Steps in a typical equivalence data analysis

Steve Simon


Categories: Blog post Tags: Hypothesis testing

There are three approaches that are used to demonstrate equivalence, each with its advantages and disadvantages.

  1. Establish adequate power
  2. Report confidence intervals
  3. Perform two one-sided tests

Each of these approaches requires that you define a range of clinical indifference first.

Defining a range of clinical indifference

[Add material to this section.]

Establishing adequate power

[Add material to this section.]

Reporting confidence intervals

[Add material to this section.]

Performing two one-sided tests

[Add material to this section.]

Modifications for non-inferiority trials

[Add material to this section.]


Assessing Equivalence: An Alternative to the Use of Difference Tests for Measuring Disparities in Vaccination Coverage. Barker LE, Luman ET, McCauley MM, Chu SY. Am. J. Epidemiol. 2002: 156(11); 1056-1061. [Abstract]

"Proving the null hypothesis" in clinical trials. Blackwelder WC. Controlled Clinical Trials 1982: 3(4); 345-53. [Medline]

Scientific and ethical issues in equivalence trials. Djulbegovic B, Clarke M. Jama 2001: 285(9); 1206-8.

Bioequivalence of generic and brand-name levothyroxine products in the treatment of hypothyroidism. Dong BJ, Hauck WW, Gambertoglio JG, Gee L, White JR, Bubp JL, Greenspan FS. Jama 1997: 277(15); 1205-13.

Statistical Significance and Clinical Relevance: A Contradiction?. Hauschke D, Statistical Solutions. Accessed on 2003-06-24.

Sample size determination for proving equivalence based on the ratio of two means for normally distributed data. Hauschke D, Kieser M, Diletti E, Burke M. Stats in Medicine 1999: 18(1); 93-105.

Choice of Control Group in Clinical Trials. International Federation of Pharmaceutical Manufacturers Associations. Accessed on 2003-02-04.

Trials to assess equivalence: the importance of rigorous methods. Jones B, Jarvis P, Lewis JA, Ebbutt AF. British Medical Journal 1996: 313(7048); 36-39. [Medline] [Full text]

Comparison of tests and sample size formulae for proving therapeutic equivalence based on the difference of binomial probabilities. Roebruck P. Statistics in Medicine 1995: 141583-94.

Of Drafts and Gales: Speaker's Forum Monthly Article. Senn S, Statistical Solutions. Accessed on 2003-11-27.

Sample size calculations for risk equivalence testing in pharmacoepidemiology. Tubert-Bitter P, Manfredi R, Lellouch J, Begaud B. Journal of Clinical Epidemiology 2000: 53(12); 1268-1274.

Equivalence trials. Ware J, Antman E. NEJM 1997: 337(16); 1159-61. [Abstract]

You can find an earlier version of this page on my original website.